Spatio-temporal patterns of active epigenetic turnover

DNA methylation is a primary layer of epigenetic modification that plays a pivotal role in the regulation of development, aging, and cancer. The concurrent activity of opposing enzymes that mediate DNA methylation and demethylation gives rise to a biochemical cycle and active turnover of DNA methylation. While the ensuing biochemical oscillations have been implicated in the regulation of cell differentiation, their functional role and spatio-temporal dynamics are, however, unknown. In this work, we demonstrate that chromatin-mediated coupling between these local biochemical cycles can lead to the emergence of phase-locked domains, regions of locally synchronized turnover activity, whose coarsening is arrested by genomic heterogeneity. We introduce a minimal model based on stochastic oscillators with constrained long-range and non-reciprocal interactions, shaped by the local chromatin organization. Through a combination of analytical theory and stochastic simulations, we predict both the degree of synchronization and the typical size of emergent phase-locked domains. We qualitatively test these predictions using single-cell sequencing data. Our results show that DNA methylation turnover exhibits surprisingly rich spatio-temporal patterns which may be used by cells to control cell differentiation.

💡 Research Summary

In this study the authors investigate how the active turnover of DNA methylation—driven by the simultaneous activity of methyltransferases (DNMT3) and demethylases (TET)—organises itself in space and time. They propose a minimal physical model that treats each CpG site as a stochastic phase oscillator whose state advances through a biochemical cycle of discrete phases. The intrinsic transition rate ω_i of each oscillator reflects local enzyme kinetics, while a coupling term k_i(ϕ) captures the influence of other sites. Because chromatin folding brings distant genomic loci into proximity, the coupling decays with genomic distance as |i‑j|^‑λ and is further limited by an exponential cutoff, ensuring a well‑defined thermodynamic limit. Importantly, the interaction is non‑reciprocal: methylated sites (ψ₂ phase) recruit DNMT3 enzymes and accelerate transitions of nearby unmethylated sites (ψ₁ phase), but the reverse effect is negligible.

By performing a system‑size expansion (Ω → ∞) of the master equation, the authors derive a Langevin equation that takes the form of a stochastic Kuramoto model with long‑range, non‑reciprocal coupling. The global degree of synchronization is quantified by the Kuramoto order parameter r(t). Coarse‑graining yields a continuity equation for the phase density ρ(θ, ω, z, t) that incorporates both the intrinsic frequency ω and an interaction‑induced drift ν̃.

First, the authors analyse the homogeneous case where all ω_i are identical. Numerical integration of the deterministic part of the Kuramoto equation shows that, starting from random phases, finite‑size phase‑locked domains emerge. Within each domain the phase differences remain constant (phase locking), while the domains coarsen over time. The coarsening speed depends non‑monotonically on the coupling strength J: weak coupling (J≈0.1) leads to slow growth, intermediate coupling (J≈1) yields the fastest coarsening, and very strong coupling (J≈10) slows growth again because large domains become internally rigid.

Next, they incorporate genomic disorder by allowing ω_i to follow an exponential distribution g(ω)=μ⁻¹e^{‑μω}. Using a strong‑disorder renormalisation group, they obtain a self‑consistent equation for the average fraction m of sites in the methylated ψ₂ phase. Analytical limits give m=½ for J→0 and m=2/3 for J→∞, while intermediate J values require numerical solution. The order parameter r is then computed from m and the interaction kernel A(λ); it never reaches unity in the thermodynamic limit, indicating that full global synchrony is impossible. Stochastic Gillespie simulations across a range of lattice sizes (N=100–1000) and phase numbers (Ω=10–100) confirm the theoretical predictions for ⟨r⟩ and ⟨m⟩ as functions of J.

Finally, the authors test key model predictions against single‑cell sequencing data that simultaneously distinguish cytosine (C), 5‑methylcytosine (5mC), and 5‑hydroxymethylcytosine (5hmC). The data reveal regional variations in methylation turnover consistent with the existence of locally synchronized domains of limited size, as the model predicts. Moreover, the observed dependence of domain size on genomic heterogeneity aligns with the theoretical role of disorder in arresting coarsening.

Overall, the paper demonstrates that chromatin‑mediated long‑range, non‑reciprocal interactions can generate phase‑locked domains of DNA methylation turnover, while genomic heterogeneity caps their growth. These spatio‑temporal patterns provide a plausible mechanism by which cells could regulate differentiation and other fate decisions through coordinated epigenetic dynamics. The work bridges molecular epigenetics with concepts from statistical physics, offering a quantitative framework for future experimental and therapeutic investigations.